Questions people ask first

What is Ḥeshbon, and what does the name mean?

Cheshbon (חשבון) is the Hebrew word for an accounting, a reckoning, the arithmetic you sit down and actually work through. That is the whole project in one word. Ḥeshbon is a data effort about how Jewish life is paid for: tuition, salaries, housing, communal giving, retirement, and the space between all of them.

Each season begins as a question I could not stop asking, gets built from datasets I had to assemble because they did not sit in any single place, and gets told as a visual essay meant to be read in a sitting. The aim across every one of them is the same, to put evidence where anecdote has been doing the work.

Who is behind it?

I am Gavriel Brown. I run a Pre-K through twelfth grade Jewish day school in Memphis, which on most days means I live inside the same arithmetic this site is about, the give and take between tuition, salaries, fundraising, and programs.

Before Memphis I taught English and started journalism programs in Newark and New York, ran curriculum in Baltimore, and served as an assistant head of school in Chicago, picking up a doctorate in education and statistics from Johns Hopkins and rabbinic ordination along the way. The About page tells the longer version as a ledger, one entry at a time.

Where do the numbers come from, and how rigorous is the method?

Every figure traces back to a primary source, most of them public records and datasets that I assembled and cleaned myself. Each story states its own sources, sample sizes, and limits, and each whitepaper ships with a methodology companion that lays out exactly how the result was produced.

My training at Johns Hopkins ran hard toward the quantitative, and the habit it left me with is to test a method rather than trust it, and to write up only what a dataset can actually support. Where a chart simplifies a published result, the story says so in plain language.

Does Ḥeshbon grade, rank, or audit individual schools or communities?

No. Nothing here audits an individual institution. The percentages describe the samples observed, several of them rest on small denominators that are noted right where they appear, and the point is always the pattern across many schools or communities rather than a verdict on any single one.

Some season and story titles are working premises rather than settled findings, and where that is the case they are labeled that way so no one mistakes a starting question for a conclusion.

How is the site organized?

The work comes in seasons, each a cluster of questions around a single theme: Jewish education, household affordability, federation and giving, and living in Israel. A season grows as its questions get answered, so the map fills in over time rather than all at once.

Within a season most findings are told twice, once as a visual essay for the reader who wants the argument in a sitting, and once as a whitepaper with a methodology companion for anyone who wants the full apparatus. A handful of interactive tools let you put your own community or your own numbers into the picture.

Who funds this, and how independent is it?

This is an independent project. I built it on my own time, no school, federation, or funder commissions the questions or reviews the findings before they go up, and nothing here is published on anyone's behalf.

Where my own vantage point could bear on a topic, for instance that I run a day school and that my wife and members of my family have worked in Jewish federations and nonprofits, I say so in the piece rather than leaving it for the reader to discover. If the funding picture ever changes, this page changes with it.

How do you use artificial intelligence in this work?

I use it the way I would a very, very productive research assistant, the kind of assistant I once was myself, except this one never needs to sleep and can work straight through the night. Early on I was the person cleaning a spreadsheet at two in the morning, chasing down one more table, reformatting a dataset by hand until it finally cooperated.

The questions are mine. The datasets are ones I chose to assemble. Every number traces back to a source I checked myself. On where the tool stops

A good deal of that groundwork now runs faster: pulling public records into a usable shape, drafting a first version of a chart, checking a calculation several different ways, putting rough language on the page that I then rewrite in my own voice. What it does not get to do is decide anything. I hold the assistant to the standard I learned to hold my own early work to, which is to test it rather than trust it, and if a number cannot be sourced and reproduced by hand it stays off the page, however confidently a model produced it.

How can I use the findings, get the data, or suggest a question?

Read freely and cite freely, and above all bring the numbers to a conversation where they help, with a spouse, a board, or a set of tuition figures in front of you. Every story names its sources so you can follow them back or check my work, which is the point of showing them.

If you have a question you cannot stop asking, a dataset I should know about, or a correction to something already published, write to me at gavriel.brown@mhafyos.org. If the material supports it, it may well turn into the next season.